I thought about Tilemap Town's scope and purpose as a project some more, and how what's already established or planned fits into that. So this isn't really new, just better organization and refining my direction a bit more?

Tilemap Town is a modern evolution of the BYOND "building game" genre with improved building, improved roleplay tools, and it borrows/adapts from Second Life where it's reasonable and cool to do so. And it's an open platform that is explicitly designed for easy client implementation, and it's supposed to be easier to access than a BYOND game.

So providing a web client (which supports touch screen devices) and a lower resource, cross-platform desktop client fits into the ease-of-access goals. Stuff like the 3DS client is additional, but lower priority access and it's justified because realistically I think it would actually get used sometimes if it got fixed, not just as a novelty thing to look at for 5 minutes. Or at least I would use it. The reusable C++ protocol implementation that came from the desktop and 3DS clients makes new client creation easier. But ideas for clients that require very old software to work, or require hardware few people have are things where I probably should acknowledge they're cool ideas but not helpful to the project, and might be a separate project.

It's fun to think about "hey, what if there was a SNES client" but if that calls for a keyboard no one has to actually be worth using then it's not super helpful. But if someone makes some addon to put a keyboard and Internet access on a device I care about (or they already have these), and I think people might actually use it if I did make a client, then sure, maybe? 3DS fits that, and DSi could too. Maybe some other handhelds could be good. It's also fun to think about "what if it was made with 2000s tech", like if there was a Tilemap Town flash client, or Java applet, and some sort of XMLHttpRequest mess. But "Tilemap Town, but with 2000s tech" already exists: it was the BYOND games I played that don't work anymore. So that's pretty clearly out of scope haha. But old computers running a modern Linux distro are supposed to be supported if possible.

And for future client/platform features that I do think fits into the project goals, I feel like it's pretty easy for stuff that makes it easier to create worlds or allows for richer worlds to be in scope, and fits with being an evolution of that genre. Like copy/paste and fills, tiles with behaviors when stepped on, more simple interactive objects, maybe simple tools for NPCs. Icon Ultima had conveyor belts and ice and doors that open when bumped into.

Same for features that make it easier to get your character into the game, like additional tools for controlling player character animation, or a way to draw them directly in the client, or perhaps even some templates.

I feel like some sort of asynchronous communication is helpful, for planning meetups if nothing else. But it's unclear what form that should take past just the simple mail feature, if anything past that is even needed. That might be something that inherently needs to be off-platform, or at least have off-platform notifications.